This research investigates the relationship among processing behaviors and their link to comprehending expository discourse. Students who orally read a chapter on geographical formations followed by a retelling. Using Pearson's chi-square and logistic regression, it was found that clauses were recalled whether the clause had been read with no miscues or with corrected miscues. Clauses containing uncorrected miscues, were significantly less likely to be retold. The sole potential predictor of main ideas was the number of clauses retold. These findings suggest that the concept of accuracy may not be as significant as suggested in the fluency research. (Contains 8 tables.)